Image identification method and image surveillance apparatus

Abstract

An image identification method is applied to an image surveillance apparatus and includes utilizing an object detecting function to generate a first bounding box and a second bounding box within a surveillance image, utilizing a foreground detecting function to generate a plurality of markers within the surveillance image, defining some of the plurality of markers which conform to the first bounding box as a first marker group and further defining other markers which do not conform to the first bounding box as a second marker group, determining whether the second marker group conforms to the second bounding box, and deciding whether the second bounding box belongs to an error of the object detecting function in accordance with a determination result.

What is claimed is: 
     
       1. An image identification method of decreasing identification error, the image identification method comprising:
 utilizing an object detecting function to generate a first bounding box and a second bounding box within a surveillance image; 
 utilizing a foreground detecting function to generate a plurality of markers within the surveillance image; 
 defining some specific markers of the plurality of markers which conform to the first bounding box as a first marker group and further defining other excluded markers which do not conform to the first bounding box as a second marker group; 
 acquiring a first acceleration of the first marker group and a second acceleration of the second marker group; 
 comparing the first acceleration with the second acceleration to determine whether the second marker group conforms to the second bounding box; and 
 deciding whether the second bounding box belongs to an error of the object detecting function in accordance with a determination result. 
 
     
     
       2. The image identification method of  claim 1 , wherein generating the first bounding box comprises:
 drawing a rectangular bounding box around a target object within the surveillance image; 
 drawing a contour of the target object; and 
 setting a maximal combined range of the rectangular bounding box and the contour as the first bounding box. 
 
     
     
       3. The image identification method of  claim 1 , wherein generating the first bounding box comprises:
 drawing a first rectangular bounding box around a target object within the surveillance image; 
 drawing a contour of the target object; and 
 generating a second rectangular bounding box by moving at least one edge of the first rectangular bounding box to align with the contour and setting the second rectangular bounding box as the first bounding box. 
 
     
     
       4. The image identification method of  claim 1 , further comprising:
 utilizing a shadow detection function to identify whether the second marker group is relevant to the first marker group; and 
 determining whether the second marker group conforms to the second bounding box in accordance with an identification result. 
 
     
     
       5. The image identification method of  claim 1 , further comprising:
 analyzing whether a vector of the first marker group is the same as or similar to a vector of the second marker group; and 
 determining whether the second marker group conforms to the second bounding box in accordance with an analysis result. 
 
     
     
       6. The image identification method of  claim 1 , further comprising:
 analyzing continuity of the second marker group relative to the first marker group; and 
 determining whether the second marker group conforms to the second bounding box in accordance with the continuity. 
 
     
     
       7. The image identification method of  claim 1 , further comprising:
 comparing a stationary period of the second bounding box with a predefined period; and 
 determining the second bounding box belongs to the error of the object detecting function when the stationary period exceeds the predefined period. 
 
     
     
       8. The image identification method of  claim 1 , wherein deciding whether the second bounding box belongs to the error of the object detecting function comprises:
 comparing a marker number of the second marker group that conforms to the second bounding box with a threshold condition; and 
 determining the second bounding box belongs to the error of the object detecting function when the marker number is smaller than the threshold condition. 
 
     
     
       9. The image identification method of  claim 1 , further comprising:
 classifying at least one marker of the first marker group that is located within the second bounding box into the second marker group when the first bounding box is partly overlapped with the second bounding box. 
 
     
     
       10. An image surveillance apparatus, comprising:
 an image receiver adapted to acquire a surveillance image; and 
 an operation processor electrically connected to the image receiver, the operation processor being adapted to utilize an object detecting function to generate a first bounding box and a second bounding box within the surveillance image, utilize a foreground detecting function to generate a plurality of markers within the surveillance image, define some specific markers of the plurality of markers which conform to the first bounding box as a first marker group and further defining other excluded markers which do not conform to the first bounding box as a second marker group, acquire a first acceleration of the first marker group and a second acceleration of the second marker group, compare the first acceleration with the second acceleration to determine whether the second marker group conforms to the second bounding box, and decide whether the second bounding box belongs to an error of the object detecting function in accordance with a determination result. 
 
     
     
       11. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to draw a rectangular bounding box around a target object within the surveillance image, draw a contour of the target object, and set a maximal combined range of the rectangular bounding box and the contour as the first bounding box. 
     
     
       12. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to draw a first rectangular bounding box around a target object within the surveillance image, draw a contour of the target object, and generate a second rectangular bounding box by moving at least one edge of the first rectangular bounding box to align with the contour and setting the second rectangular bounding box as the first bounding box. 
     
     
       13. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to utilize a shadow detection function to identify whether the second marker group is relevant to the first marker group, and determine whether the second marker group conforms to the second bounding box in accordance with an identification result. 
     
     
       14. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to analyze whether a vector of the first marker group is the same as or similar to a vector of the second marker group, and determine whether the second marker group conforms to the second bounding box in accordance with an analysis result. 
     
     
       15. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to analyze continuity of the second marker group relative to the first marker group, and determine whether the second marker group conforms to the second bounding box in accordance with the continuity. 
     
     
       16. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to compare a stationary period of the second bounding box with a predefined period, and determine the second bounding box belongs to the error of the object detecting function when the stationary period exceeds the predefined period. 
     
     
       17. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to compare a marker number of the second marker group that conforms to the second bounding box with a threshold condition, and determine the second bounding box belongs to the error of the object detecting function when the marker number is smaller than the threshold condition. 
     
     
       18. The image surveillance apparatus of  claim 10 , wherein the operation processor is further adapted to classify at least one marker of the first marker group that is located within the second bounding box into the second marker group when the first bounding box is partly overlapped with the second bounding box.